Slam algorithm for aruco landmark array based on synchronization optimization

Abstract

High precision mapping algorithm is the core of two-dimensional(2D) landmark array(the ArUco landmark is used in this paper) localization application. In most of the existing mapping algorithms, the accumulated errors of online methods can not be eliminated, meanwhile, the manual calibration and off-line methods are inefficient and time-consuming. In order to overcome those shortcomings, a real-time simultaneous localization and mapping(SLAM) algorithm for landmark array based on synchronized optimization is proposed in this paper. The proposed algorithm updates and corrects the global map online, and the accumulated errors is optimized based on minimizing the re-projection error of the landmarks ' vertexes. In order to verify the proposed algorithm, several contrast experiments with other mapping systems are presented. The experimental results show that the proposed algorithm has the advantages of high map precision and small accumulative error.